Analyzing functional similarity of protein sequences with discrete wavelet transform

Area of Science:

• Bioinformatics and Computational Biology
• Structural Bioinformatics
• Genomics and Proteomics

Background:

• Identifying functional similarity in proteins with low sequence identity is a significant challenge in bioinformatics.
• Existing methods like pairwise alignment and PSI-BLAST have limitations in detecting distant homology.
• Novel computational approaches are needed to accurately infer protein function from sequence data.

Purpose of the Study:

• To develop and validate a new method for detecting functional similarity between protein sequences.
• To address the challenge of low sequence identity in homology detection.
• To introduce a robust metric for quantifying pair-wise protein similarity.

Main Methods:

• Application of discrete wavelet transform (DWT) on protein sequences.
• Integration of various protein substitution models with DWT.
• Development of a novel 'S' function metric for pair-wise similarity assessment.
• Implementation of a DWT-based segmentation technique for long protein sequences.

Main Results:

• The proposed DWT-based 'S' function effectively measures pair-wise protein similarity.
• The method demonstrates capability in identifying functional similarity for proteins with low sequence identity.
• The segmentation technique successfully handles long protein sequences.

Conclusions:

• The discrete wavelet transform offers a powerful approach for functional protein similarity detection.
• The novel 'S' function provides a reliable metric for assessing pair-wise protein relationships.
• This DWT-based methodology complements and potentially surpasses traditional alignment methods for specific homology detection tasks.